Chapter 9 Flashcards
What is schematic knowledge?
General background knowledge through experience
How do we store knowledge?
1) we use regularity (consistency in the way different people store knowledge)
2) we use previous knowledge to make inferences
3) much of our knowledge is implicit (common sense)
What else is much of our knowledge?
Semantic memory (facts)
What is a category?
Set of items that are similar
Ex. Cups
What is an exemplar?
Individual items in a category
What is a concept?
Mental representation of objects
Idea that includes all characteristics associated with it
What is the common sense knowledge problem (AI)?
Problem in classic AI in which computers don’t possess that same commons sense knowledge as humans, cuz knowledge has to be explicit in classic AI compared to humans ability to infer implicit knowledge.
What does the classical approach to categorization use for categorization?
Defining features (feature that is necessary for defining category membership)
What is the issue with using defining features in the classical approach?
Many things we classify in categories are missing some defining features
What are typicality effects?
We tend to rate some animals as more typical exemplars then others
Ex. Bird:
Robin vs ostrich
We will behave different to typical vs atypical items
What is semantic priming?
We respond faster when shown a similar word before
If I wrote a sentence about a turkey (atypical) and a robin (typical) would I be more likely to recall the sentence then if opposite order (typical then atypical?)
No.
We are more likely to recall sentences when the first item is typical and the second is atypical
Vs
When first is atypical and the second is typical
What is the prototype theory of categorization?
We consider the most characteristic feature (feature most likely possessed but not required) of a category when we encounter exemplars. (Fuzzy boundaries)
Stored in a prototype (mental average of category members)
-these don’t exist in reality
-explain typically effects based on amount of shared features with prototype
Category members have family resemblance (items can belong in one category as long as one shares a feature with one another)
Does context impact our prototype (give an example)
Yes. Ex an Australian might see different bird more often
Maybe ostrich is typical for them
What is the exemplar theory of categorization?
We stores actual examples we encountered in the past
Categorization happens by comparing new items to the ones in your memory
Can also explain context, because this depends on individual experience
What is the knowledge based theories of categorization ?
We use more than feature similarity. We use our broad knowledge base to explain category membership.
Ex. Dogs are doggy
What is physiological essentialism?
Proposal that categories have a natural and underlying true nature that cannot be stated, explicitly
We rely on categories to teach us about the world
What are basic level categories? (Lvl ppl find most natural)
The most cognitively efficient version
Example is dog
Category first learned by children
What is subordinate categories?
Below the basic level more informative, but less distinctive
For example, a German Shepherd, or poodle
What is the super ordinate category level?
Above basic level, less informative than basic, but more distinct
For example, animal .
What are semantic network models?
Knowledge is stored as concepts with an individual units called nodes
They point to properties of those concepts as well
Model shows cognitive economy (conserve effort)
And property inheritance (hierarchy of properties)
What is the spreading activation model?
Semantic network model in which concepts are organized, based on semantic similarity to each other
There’s no hierarchy in this one
Could explain typicality because we jump around mentally based on loosely connecting characteristics
What is the schema of view of knowledge?
Knowledge is organized into groups of related schemata (plural of schema)
Made method of repeated reproduction
-If you’re shown a stimulus once
Asked to redraw it over and over after a delay
It will become less and less like a background image in more like a familiar object
What is the symbol grounding problem in AI?
How symbols get their meaning in the real world. Need a way to connect to the real world to avoid an endless cycle of symbolic representation.
